Foundations of Software Engineering

Size: px

Start display at page:

Download "Foundations of Software Engineering"

Dominick Wheeler
6 years ago
Views:

1 Foundations of Software Engineering Part 2: Quick intro to process, teamwork, risk and scheduling Christian Kästner 1

2 Learning Goals Recognize the Importance of process Understand the difficulty of measuring progress Identify what why software development has project characteristics Use milestones for planning and progress measurement Ability to divide work and planning and replan it Model dependencies and schedule work with network plans and Gantt diagrams Identifying and managing risks 2

3 About You Open Source Saw project fail, know why yes no yes no Frustrating team experience multiple times no once 3

4 Process 4

5 How to develop software? 1. Discuss the software that needs to be written 2. Write some code 3. Test the code to identify the defects 4. Debug to find causes of defects 5. Fix the defects 6. If not done, return to step 1 5

6 What is a Software Process? 6

7 Software Process The set of activities and associated results that produce a software product Sommerville, SE, ed. 8 7

8 Example of Process Decisions Writing down all requirements Require approval for all changes to requirements Use version control for all changes Track all reported bugs Review requirements and code Break down development into smaller tasks and schedule and monitor them Planning and conducting quality assurance Have daily status meetings Use Docker containers to push code between developers and operation 8

9 100% Percent of Effort 9 0% Project beginning Time Project end

10 100% Trashing / Rework Percent of Effort Productive Coding 0% 10 Project beginning Time Project end

11 100% Trashing / Rework Percent of Effort Productive Coding Process: Cost and Time estimates, Writing Requirements, Design, Change Management, Quality Assurance Plan, Development and Integration Plan 0% 11 Project beginning Time Project end

12 100% Trashing / Rework Percent of Effort Productive Coding Process 0% 12 Project beginning Time Project end

13 Example process issues Change Control: Mid-project informal agreement to changes suggested by customer or manager. Project scope expands 25-50% Quality Assurance: Late detection of requirements and design issues. Test-debug-reimplement cycle limits development of new features. Release with known defects. Defect Tracking: Bug reports collected informally, forgotten System Integration: Integration of independently developed components at the very end of the project. Interfaces out of sync. Source Code Control: Accidentally overwritten changes, lost work. Scheduling: When project is behind, developers are asked weekly for new estimates. 13

14 Survival Mode Missed deadlines -> "solo development mode" to meet own deadlines Ignore integration work Stop interacting with testers, technical writers, managers, 14

15 100% Trashing / Rework Percent of Effort Productive Coding 15 0% Project beginning Process Time Project end

16 Hypothesis Process increases flexibility and efficiency Upfront investment for later greater returns 16

17 17

18 Process models Ad-hoc Waterfall Spiral Agile More later 18

19 19

20 Estimating Effort 20

21 Task: Estimate Time A: Java version of the Monopoly boardgame with Pittsburgh street names (you) B: Bank smartphone app (you with team of 4 developers, one experienced with iphone apps, one with background in security) Estimate in 8h days (20 work days in a month, 220 per year) 21

22 Revise Time Estimate C: Remember Scrabble experience. Is Monopoly similar/different/easier/more challenging/reusable? How much design did you do? Break down the task into ~5 smaller tasks and estimate them. Revise your overall estimate if necessary 22

23 23

24 Constructive Cost Model (Cocomo) Regression formula based on project history Requires experience with similar projects Encourages documentation of experience 24

25 Constructive Cost Model (Cocomo) Ratings Cost Drivers Very Low Low Nominal High Very High Extra High Product attributes Required software reliability Size of application database Complexity of the product Hardware attributes Run-time performance constraints Memory constraints Volatility of the virtual machine environment Required turnabout time Personnel attributes Analyst capability Applications experience Software engineer capability Virtual machine experience Programming language experience Project attributes Application of software engineering methods Use of software tools Required development schedule Sommerville, SE, Chapter 26

26 Study: Variability and Reproducibility in Software Engineering Study by Simula Research Lab in Norway Created System Requirements Specification for a web information system (11 pages) Received bids from 35 companies, 14 with schedule Contracted 4 companies to build the same system

27 Bids and time estimations No relationship between price and planned time or methods in bids

28 Anda, Bente CD, Dag IK Sjøberg, and Audris Mockus. "Variability and reproducibility in software engineering: A study of four companies that developed the same system." IEEE Transactions on Software Engineering 35.3 (2009):

29 Development Process

30 Risk and Uncertainty 30

31 Risk management Key task of a project manager Identify and evaluate risks early If necessary, plan mitigation strategies Document results of risk analysis in project plan Project risks: scheduling and resources e.g., staff illness/turnover Product risks: Quality and functionality of the product e.g. used component too slow Business risks: e.g., competitor introduces similar product 31

32 32

33 Sources of Uncertainty Unpredictable operating environment Cybersecurity threats, device drivers Unanticipated usage scenarios Limited predictive power of models Halting, abstract interpretation, testing Bounded rationality of humans Designers, developers Customers, users 33

34 Innovative vs Routine Projects Most software projects are innovative Google, Amazon, Ebay, Netflix Vehicles and robotics Language processing, Graphics Routine (now, not 10 years ago) E-commerce websites? Many control systems? Routine gets automated -> innovation cycle 34

35 Case 1 35

36 Case 2 36

37 37

38 Discussion What are important risks in early and late phases of development? Analyze risks to rank them How can they be managed/mitigated? 38

39 Accepting and Coping with Risks Selectively innovate to increase value Improve capability and competitiveness Focus risk where it is needed Rely on precedent and convention (experience) Use iteration and feedback prototypes, spiral development, sprints 39

40 Managing Risks Address risk early Prototyping, spiral development Identify mitigation strategies 40

41 Example: Bank Software Build on vendor and open-source components OS, DB, app server, tooling, backups, information management, Stick with reliable vendors and integrators Focus internal risk-taking on business differentiators Web-site features, internal queries, decision algorithms Outsource to access expertise, lower costs, gain flexibility Black-box testing Overall system architecture Internal engineering practices and tools 41

42 Risk analysis Estimate likelyhood and consequences Requires experienced project lead Rough estimations usually sufficient very low (<10%), low (<25%),... catastrophic, severe, acceptable, neglegible Focus on top 10 risks 42

43 Risk planning Risk avoidance Risk minimization Emergency plans 43

44 Strategies to help manage risk Risk Organizational financial problems Recruitment problems Staff illness Defective components Requirements changes Organizational restructuring Database performance Underestimated development time Strategy Prepare a briefing document for senior management showing how the project is making a very important contribution to the goals of the business and presenting reasons why cuts to the project budget would not be cost-effective. Alert customer to potential difficulties and the possibility of delays; investigate buying-in components. Reorganize team so that there is more overlap of work and people therefore understand each other s jobs. Replace potentially defective components with bought-in components of known reliability. Derive traceability information to assess requirements change impact; maximize information hiding in the design. Prepare a briefing document for senior management showing how the project is making a very important contribution to the goals of the business. Investigate the possibility of buying a higher-performance database. Investigate buying-in components; investigate use of a program generator.

45 Planning 45

46 (Software) Projects One-time endeavour; unique wrt Goals time, financial, personal, and other constraints differences to other endeavours project-specific organization Defined, limited time (clear start and end) Clear goals Constraints: budget, resources,... Requires management high risk coordination of experts Software development always proceeds in projects New/unique goals, innovatice technology Intangable result, progress hard to measure Software projects tend to fail more often than industrial projects 46

47 Measuring Progress? I m almost done with the app. The frontend is almost fully implemented. The backend is fully finished except for the one stupid bug that keeps crashing the server. I only need to find the one stupid bug, but that can probably be done in an afternoon. We should be ready to release next week. 47

48 Measuring Progress? Developer judgment: x% done Lines of code? Functionality? Quality? 48

50 Milestones and deliverables Making progress observable, especially for software Milestone: clear end point of a (sub)tasks For project manager Reports, prototypes, completed subprojects "80% done" not a suitable mile stone Deliverable: Result for customer Similar to mile stone, but for customers Reports, prototypes, completed subsystems 50

51 Case 3 Define tasks and milestones for the Monopoly implementation Which tasks have dependencies? 51

52 52 xkcd.com/518

no Reestimate project parameter yes new feature requests Create schedule no

53 Project Planning Identify constraints Estimate project parameters Define milestones Budget, Personal, Deadlines activities begin Check progress Done? no Reestimate project parameter yes new feature requests Create schedule no Refine schedule Problem? yes Abort? renegotiate constraints Technical review 53

54 Project Manager Tasks Divide project into work packages with measurable outcomes (each 1-10 weeks) Estimate time and resources Create order and determine parallelism Plan buffers for anticipated and unanticipated problems Software available, such as, Microsoft Project, GanttProject, Kplato, etc Requires experience for estimation 54

55 Activity Network Work package Milestone 55

56 Gantt Diagrams 56

57 Replanning Inaccurate time predictions are normal -> Update schedule Task Planned Actual Difficulty/Risk Responsibility Completed 57 Einführung in die Softwaretechnik

58 Reasons for Missed Deadlines Insufficient staff (illnesses, staff turnover,...) Insufficient qualitication Unanticipated difficulties Unrealistic time estimations Unanticipated dependencies Changing requirements, additional requirements Especially in student projects Underestimated time for learning technologies Uneven work distribution 58

59 Recognize Scheduling Issues Early Monitoring and formal reporting necessary Establish who, when, what Compare planned/actual data Measurable milestones Outdated schedules no meaningful management mechanism 59

60 % completed 90% 100% Almost Done Problem Last 10% of work -> 40% of time (or 20/80) planned actual Make progress measureable Avoid depending entirely on developer estimations reported progress 60 time

61 Milestone Trend Analysis Actual time Estimated completion time 61

62 Milestone Trend Analysis Actual time Estimated completion time Quickly rising estimations too optimistic Changing trends unreliable early estimations Ziz-zag pattern unreliable estimations Falling overly large buffers 62

63 Mitigation strategies Additional personal especially experts for specific tasks Temporarily increasing hours of work (overtime, vacation freeze) short-term solution only Improve tooling, methods, and processes Buy, contract, off shore Renegotiate / reduce functionality Set priorities, incremental deployment Move deadline Avoid: less testing/quality assurance 63

64 Team productivity Brook's law: Adding people to a late software project makes it later. 64

65 Teamwork (Student Teams) (more on teams in real projects later in the course) 65

66 Expectations Meet initially and then regularly Review team policy Divide work and integrate Establish a process Set and document clear responsibilities and expectations Possible Roles: Coordinator, Scribe, Checker, Monitor Rotate roles every assignment Every team member should understand the entire solution 66

67 Team Policies see document Make agreements explicit and transparent Most teams will encounter some problem 67

68 Dealing with problems Openly report even minor team issues in individual part of assignments In-class discussions and case studies Additional material throughout semester We will attend one team meeting 68

69 Planning and In-Team Communication Asana, Trello, Microsoft Project, Github Wiki, Google docs, , Slack, Facebook groups, 69

70 Homework and Readings 70

71 Reading assignment Prepare a case discussion about Healthcare.gov Read the TIME Magazine article Code Red Explore Wikipedia, Youtube, other resources Look at the webpage itself Think about what went wrong and why (both technical and nontechnical issues), how it could be fixed or prevented in future projects, 71

72 Homework 1 Team assignment Build grad school application system Process focus Estimate and track time Plan process and observe deviations Analyze risks 72

73 73

74 Further Reading McConnell. Software Project Survival Guide. Microsoft Press 1998, Chapter 3 Sommerville. Software Engineering. 8 th Edition. Addison-Wesley Chapters 5 "Project Planning" and 26 "Software Cost Estimation" 74

Foundations of Software Engineering

Foundations of Software Engineering Part 2: Quick intro to process, teamwork, risk and scheduling Christian Kästner 1 Learning Goals Recognize the Importance of process Understand the difficulty of measuring